Chapter 4 — Prescription, Dosing & Fluid Kinetics

A CRRT prescription is four numbers and one architectural choice. Getting the numbers wrong is usually recoverable; getting the architecture wrong quietly halves the delivered dose.

4.1 The Components of the Prescription

Effluent = Qd + Qr + Net UF. This total, indexed to body weight, is the dose.

💡 The Qb misconception. Trainees routinely raise Qb hoping to raise clearance. In CRRT, Qd and Qr are so far below Qb that the dialysate/filtrate is saturated regardless. Raising Qb from 150 to 250 mL/min lowers filtration fraction and extends filter life — both valuable — but does not raise urea clearance. Dose is set by effluent, not by blood flow.

4.2 The Dose Trials

ATN / VA-NIH Acute Renal Failure Trial (NEJM 2008)

• Population: 1,124 critically ill patients with AKI requiring RRT, with at least one non-renal organ failure or sepsis.
• Design: modality was assigned by haemodynamic status, and intensity was randomised within it.
• Intensive: IHD or SLED six times weekly, or CVVHDF at 35 mL/kg/h.
• Less intensive: IHD or SLED three times weekly, or CVVHDF at 20 mL/kg/h.
• Primary outcome: 60-day all-cause mortality.
• Result: No difference — 53.6% intensive vs 51.5% less intensive (p = 0.47). No difference in recovery of kidney function or non-renal organ failure. Hypotension during treatment was more frequent in the intensive arm.

RENAL (NEJM 2009)

• Population: 1,508 critically ill patients with severe AKI, in Australia and New Zealand.
• Design: CVVHDF in all patients, randomised to effluent of 40 mL/kg/h vs 25 mL/kg/h.
• Primary outcome: 90-day all-cause mortality.
• Result: No difference — 44.7% in both arms. No difference in RRT dependence among survivors.
• Key safety finding: hypophosphataemia was significantly more common in the higher-intensity arm (approximately 65% vs 54%).

Synthesis

✅ The conclusion is genuinely settled. Two large, well-conducted trials with different designs, different modalities and different outcome timepoints converge: increasing effluent dose above roughly 20–25 mL/kg/h does not improve survival, and does cause measurable harm. This is one of the firmest findings in the CRRT literature.

4.3 Guideline Anchoring

KDIGO 2012 AKI, Section 5.8:

• Deliver an effluent volume of 20–25 mL/kg/h for CRRT in AKI. (1A)

• This will usually require a higher prescription of effluent volume. (2B)

• The dose of RRT to be delivered should be prescribed before starting each session, and delivered dose should be assessed frequently to adjust the prescription. (1B)

• For intermittent and extended RRT in AKI, deliver a Kt/V of 3.9 per week. (1A)

💡 The 1A/2B pairing is the most operationally important sentence in KDIGO's RRT section. The delivered dose is the graded recommendation; the prescribed dose must exceed it. In practice, delivered dose falls short of prescribed by a meaningful margin because of downtime — filter clotting, circuit changes, imaging, theatre, procedures. Prescribing exactly 25 mL/kg/h reliably under-delivers.

The weight question

⚠️ Declared gap: KDIGO does not specify which body weight to index dose to, and no trial has compared actual, ideal, or adjusted body weight as the dosing denominator. RENAL and ATN used actual body weight at enrolment. In marked obesity, indexing to actual weight can prescribe very high absolute effluent volumes with no evidence of benefit and a plausible increase in the harms seen in the high-dose arms. Document which weight was used and why; be aware the choice is unvalidated either way.

4.4 Pre-Dilution vs Post-Dilution — The Architectural Choice

Where replacement fluid enters the circuit changes both clearance and circuit life, in opposite directions.

Post-dilution

Replacement fluid enters after the filter. Blood arrives at the membrane at full plasma solute concentration.

• Advantage: maximal solute clearance per litre of effluent. Convective clearance ≈ Q_uf × Sc, with no correction needed.
• Disadvantage: all ultrafiltration is drawn from undiluted plasma, so filtration fraction rises steeply. Progressive haemoconcentration along the fibre → protein layering → clotting.
• Practical ceiling: post-dilution alone constrains achievable dose, because FF must stay below roughly 20–25%.

Pre-dilution

Replacement fluid enters before the filter. Blood is diluted on arrival at the membrane.

• Advantage: lowers haematocrit and viscosity entering the filter, reduces filtration fraction, and prolongs circuit life. Also permits higher total ultrafiltration rates.
• Disadvantage: the hemodilution effect — solute concentration presented to the membrane is reduced, so each litre of effluent carries less solute. Delivered dose is lower than the effluent volume suggests.

The dilution correction

Correction factor = Qb_plasma / (Qb_plasma + Q_pre)

Effective dose = prescribed effluent × correction factor

A worked illustration: with plasma water flow of roughly 2,000 mL/h entering the filter and 2,000 mL/h of pre-dilution replacement fluid, the correction factor is approximately 0.5 — the delivered small-solute dose is around half the nominal effluent volume. This is the single most common cause of silent under-dosing in CRRT.

💡 Clinical pearl: Many modern machines display effluent dose without applying the pre-dilution correction. If your prescription is heavily pre-dilutional, the number on the screen is an overestimate of delivered small-solute clearance. Know whether your device corrects, and by what formula.

Comparison

⚠️ Declared gap: although the kinetic trade-off is well characterised and reproducible, no randomised trial has shown that pre- versus post-dilution affects patient survival. Trials and observational data have examined filter life and clearance efficiency — surrogate endpoints. Choose on circuit logistics, and correct the dose arithmetic; do not claim an outcome benefit.

Mixed dilution — splitting replacement fluid between both ports — is widely used to capture part of each advantage. The split should be chosen to hold FF in range while keeping the correction factor acceptable.

4.5 Fluid Kinetics: Net Ultrafiltration

Net UF is prescribed independently of dose and is the only component that changes the patient's volume status.

⚠️ Declared gap — an active area of uncertainty. Observational data have associated both very high and very low net ultrafiltration rates with worse outcomes, generating interest in an intermediate "safe" range. These are observational associations, subject to confounding by indication, and have not been confirmed by randomised trials. There is no validated net UF rate. Titrate to haemodynamics and fluid balance goals, reassess frequently, and do not treat any published rate as a target.

4.6 A Worked Prescription Logic

1. Set the dose target: aim for a delivered 20–25 mL/kg/h. Prescribe above it to absorb downtime.
2. Choose the modality (Ch. 2) — this determines whether the dose is delivered as Qd, Qr, or both.
3. Set Qb to hold filtration fraction below ~20–25% at the intended Q_uf, not to chase clearance.
4. Choose the dilution architecture, then apply the correction factor to confirm the delivered dose still meets target.
5. Set net UF separately, to haemodynamics and fluid balance — never bundled into the dose calculation.
6. Audit delivered dose daily. KDIGO grades this 1B. Downtime is where prescriptions go to die.

4.7 Chapter Summary

1. Effluent = Qd + Qr + net UF. That total, per kg per hour, is the dose.
2. ATN and RENAL agree: more is not better. Higher intensity produced no survival benefit and more harm (hypotension; hypophosphataemia).
3. KDIGO: deliver 20–25 mL/kg/h (1A); prescribe higher (2B). The gap between the two is downtime.
4. Qb sets filtration fraction, not dose.
5. Pre-dilution buys circuit life and pays for it in clearance. Apply the correction factor or you are under-dosing without knowing it.